The reason I’m expressing myself in such detail today is twofold. First, it’s always important to hear all sides of a story and objectively assess and evaluate what you’ve heard. Second, the knowledge gained, as of today, is now completely different. This isn’t to excuse what could have been done earlier but wasn’t. And this isn’t just about crisis management, because if that becomes necessary, something has already gone significantly wrong beforehand. Anger and disappointment (including with oneself) are, in today’s media-saturated age, the worst advisors. Three experts, five opinions. In the end, the lawyer’s advice was the most sensible. Therefore, today’s text is without prior discussion and lengthy correction, but with the necessary distance. And I must also comment on two videos that really shouldn’t have been necessary.
Preliminary Validation
Please read the following paragraph carefully, even though it does not meet the standards of a robust review. I currently lack the nerves for that:
I replicated Roman’s test with a 120mm radiator and a Core i9-13900K on a test system, where I replaced the original radiator with a smaller NexXxos ST 30, and got very similar results to Roman. Using the poorer NF-A12x25 PWM from 2020, both fans are roughly equal at 2000 RPM, and there is even partial thermal throttling. The 2023 model beats the Apex by 2 to 3 degrees at the same RPMs, staying just below the limit depending on power loss and room temperature. Below 150 watts, the results become relative, and one would have to measure more consistently with more water and over a longer period. This isn’t possible right now, but I am following third-party tests with due attention.
Thus, the Alphacool Apex Stealth Metal Fan is still a usable fan, but in practice, it’s far from what was promised. For this test with a few fans and only two runs per fan, I spent half the night to complete this statement with my interim result. Also because I felt spurred by the “sitting out” approach to at least do something at all.
What remains in the end are great looks and feel, as well as good decoupling on an otherwise mediocre product. Nothing more. As a case fan, with a diffuser or shroud, it’s certainly usable, but definitely worlds away from what was hyped up, partly through my own mistakes. I stand by this and must take full responsibility at this point. No ifs or buts. But it was important for me to replicate this on-the-fly to provide a practical basis for what follows.
The Role of Freelancers and Oversight Lapses
The website bears my name, but is now shaped by several people. This is good and due to the versatility, even with niche topics, it’s absolutely beneficial. This means no permanent employees or contractually bound freelancers. Everything is based on a voluntary principle where someone contributes only when they have the time and inclination. Therefore, I avoid too much influence on product selection to allow creativity to flourish, but I make sure no companies unfairly gain a foothold through this back door. I even rotate topics and tasks from time to time.
My first mistake, and I must admit this, was allowing a freelancer to complete tests for third parties on our equipment in his free time without maintaining control myself. This kind of side job is free for everyone here, as long as it’s taxed, but I should have defined the rules much tighter and clearer. This includes being able to access the results obtained at any time. This is also necessary to assess whether the whole thing makes sense in terms of content and expertise. I will and must handle this completely differently in the future, not only with fans. This naturally includes the use of my name and logo. For these reasons, I have strictly outsourced all these things that I personally do for third parties to a second company and never tested on the website products that I actively co-developed or publicly reviewed products from companies with whom I have a consulting contract or other business relationships. This must be strictly enforced, under penalty of one’s own downfall, and it’s also an unconditional obligation to the readers.
The second mistake was not giving the freelancer clear contractual rules about such things and dealing with third parties. However, one needs these to some extent to protect him from himself right from the start. Yes, through such activities one naturally gets closer contact with all sorts of industry representatives and perhaps even builds a more personal relationship with some on a private level. But at this point, one either needs the personal maturity of many years to cleanly separate between different concerns or permanent control to not violate usual transparency rules. I neglected and hardly controlled this important point, unfortunately. That this is now backfiring should have been foreseeable.
Background to the Fan Database
This brings us to the first time to the measuring setup and Pascal’s role. The development of the technology behind it is a long story because Pascal did quite good practical work here, creating an interesting solution under the direction of a cooling device manufacturer, which Pascal then implemented decently according to his possibilities. It was calibrated and tested at the time by the manufacturer, where Pascal also received training. The history and the measuring setup as such are not “secret”, but have been completely documented in the respective articles on the topic up to the fan database. Moreover, the methodology and the desired result were discussed for months in the forum and implemented step by step by us in terms of content. At this point, I made the mistake of relying blindly on the “public joker” forum and the majority of idea contributors regarding content and its interpretation. Unfortunately, without realizing in which direction one might be heading. One can completely over-optimize something and not even notice it. I definitely have to take this shoe on today’s view. As a measuring device freak, one quickly gets infected by this virus.
The combination of theory and practice should have been urgently advised back then, because one could have quickly noticed if measurement results do not correlate with the real values achievable in actual installation. Belief in calibrated technology can also blind, and the ubiquitous race for ever better values can quickly detach one from reality. But I had already written something about this point in the first statement and also announced the corresponding steps regarding future tests. All this really takes time, which we just don’t have right now. Nevertheless, I’m taking it with the support of uninvolved third parties, which definitely has nothing to do with an alleged “sitting out”. I even have to, because the cause research is immensely important. Ideally, a mix of theory and practice. Let’s call it a learning curve, I’m happy to be taught about it. But more on that later.
The Creation of the Article on the Alphacool Apex Stealth Metal Fan
This brings us to the fan test of the Apex and my part. End of September, beginning of October, I’m not going to pin myself down now, the final values of the prototype were in the fan database. Only visible if you were logged in or knew the hidden link. At this time, I was also asked by Alphacool if they could use the measurement values for their own marketing. I found this somewhat strange, considering that the tests of the prototypes were apparently paid for by the freelancer from my perspective. I don’t know either the invoices or the extent, but it probably wasn’t even that much. However, a few things contradict each other now. According to Alphacool, no more invoices were issued for the last tests and the messenger histories after October 4, 2023, were unfortunately almost completely deleted by Pascal.
The fact that my logo later appeared in the partially even erroneously reconstructed chart graphics of the marketing material is all the more annoying because it implies a business relationship with outsiders that never existed. My mistake at the time was not to strictly prevent this or at least to inquire. But once again: There was and is no contract between Alphacool and me about these tests. That I more or less silently tolerated it is also my sole omission and certainly one reason for the damage incurred.
Then came the offer to publish the test exclusively and shortly before everyone else. The circumstances seemed good because the manufacturer assured me that other editorial teams would receive (have received) these pre-series fans and one could read their tests publicly in two to three weeks, which would then have been associated with a market launch. Alphacool also confirmed to me upon request that the Apex in their own tests on a radiator had been significantly quieter than the Noctua at the same performance. That I relied on this and only did a short candle test, where you check the distance to blowing out a candle, was also a leap of faith in hindsight that backfired because it was not justified.
Later, after the article was published and I asked where the other reviews were, I learned for the first time that the manufacturer had recalled all other prototypes from the reviewers. The explanation that the test results could turn out too poor due to misplaced inserts during transport seemed quite plausible to me, because the fiddling with changing the metal shells on my own prototype was excruciating. I almost needed an hour to center it precisely on my copy, for which I even had to build a tool first. I had, at that time (you can also see it in the photos in the article), all possible iterations of the fan, but only with a 4-pin motor. There was actually nothing to counter-test.
The review should have been either taken offline or thoroughly revised by that time, especially in regards to the awards, conclusion, and headline. This is something I can be rightly criticized for. Particularly in light of the pre-order that was launched because the manufacturer realized that the screw idea was not practical. Changing the production obviously took more time, but I should have been consistent from that point on. I only retroactively marked the database entries, which are displayed in the article via a shortcut, as prototypes, which was definitely not sufficient. The article was published under my name simply because I wrote it finally and also did all the material analyses. One might think that without this, I would have been off the hook, but unfortunately, it’s much more complex.
To counter suspicions that it was a paid “fake,” Pascal made a live cell phone video for me, which showed the measurement without cuts, from the settings to the measurement and the values of around 189 m²/h. This saved me a 12-hour drive back and forth. This video can be viewed again in the forum thread. However, if I had been aware of all the trouble about his job and the still existing uncertainties in this form, the review would certainly not have happened. At least not without big question marks. Because it actually gets much worse.
Today I know, both from Alphacool and from third parties, that the manufacturer might have already been aware of the fan’s weaknesses at that time. And with that, we come to the next statements, where Pascal and Alphacool contradict each other. According to Alphacool, which I also believe, Pascal received measurements from the OEM and also other material. I have not seen any of this from him to this day. It can no longer be verified, because this communication, unfortunately, no longer exists.
The fact that customers were misinformed and lured into pre-orders is the result of a pre-order marketing that Alphacool approached far too enthusiastically and is certainly to blame for at first. But my article was still online and contributed to this. I could have even ended it with a bang at that point, even should have, but for that, I would have needed a final fan or the last prototype for testing, which I didn’t have. And various documents, which Alphacool probably assumed I had. That doesn’t make it better, but when in doubt, it’s better not to test anything than to do something like this.
Yes, I should have just done it, I know, even if it’s by listing all the question marks as an official reason. But even there, I still trusted the statements made to me, saying that they had achieved very similar results themselves. Maybe Alphacool will write something about it, because it’s not up to me to write about things I can’t verify. I was promised something like a separate statement for the first response on Saturday, but then it didn’t happen.
When the final samples finally arrived in December, I should have at least tested them under practical conditions for a short time. Roman did that, and we had more than one long phone conversation. There’s nothing to criticize about his tests (except maybe the sound level measurement), which is why I didn’t test anything big myself. A test to the test would be pointless if it yields the same result, where everything has already been said, and also, due to important private reasons, I simply couldn’t do it until New Year’s. I have to politely insist on this, because there are much more important things in private life than something like fans. I ask for understanding for this.
However, I have, as already stated in the first statement, immediately arranged for the measuring setup to be picked up from Pascal by an uninvolved third party and transparently checked elsewhere. After that, plausible and reliable retests with the inclusion of practical counter-tests are planned. More will be written about this. But even yesterday I was still of the opinion that this has to be implemented by uninvolved third parties with the necessary care and without counterproductive haste. Neither Pascal nor I may be actively involved here. That would be like sending a stork to a pond to count frogs. Transparency starts with me when you completely take yourself out of the equation. Therefore, the term “sitting it out” is a formulation with which I do not conform. Especially since Fritz, the appointed examiner, has also spoken with Roman for a long time. To now qualify these efforts with a single, rather negatively connoted word is at least not very collegial. Not even for the colleague who will accompany this error analysis and reprocessing qualified and later also conduct the tests.
Pascal’s Role and Crisis Management
This is the part where I needed the most distance. However, I must also be careful not to let too many personal feelings flow into it, because it was also a kind of self-help project over the entire time. It’s not the first case where I try to help people get back on their feet with my reach and technical possibilities, who, in my opinion, deserve it but can’t do it alone. This is lived community for me, and I would do it again anytime. Just completely differently now.
Without going into detail, but some (also personal) incidents with Pascal last week completely threw me off track. That must suffice. And so, in my statement on Saturday, I initially formulated some things very emotionally and then let myself be carried away to statements in the forum, which I better should not have made. After comments and PMs came from several sides that one shouldn’t do it like this and it would be better to remove everything, instead of discussing it for hours, I did exactly that. Especially since I have completely lacked the strength to think somewhat freely for almost a week, let alone work productively at all. By the way, the forum thread on the statement has not been deleted, but was only set to private. I have now reactivated this thread after a very long conversation with Pascal.
This also applies to the news about Pascal’s personnel change to Alphacool, which was temporarily set to private on Saturday after multiple advisements, to first find out all the backgrounds. Because be it as it may, I am very reluctant to make myself available for such things.
Pascal, who worked here as a contractually unbound freelancer (i.e., without an exclusive commitment), communicated to me at the beginning of October that he finally got a permanent job at Alphacool or had already accepted it, which will occupy him all day. This means a “full-time job” in my understanding. At first, I was even a bit disappointed, because I see something like this also as a kind of “poaching”. However, after some reflection, I was also fair enough to admit to him that a real full-time job is more valuable than a freelancer story, where you top up your social benefits and pursue your hobby.
Regarding what he was supposed to do there exactly, however, there are two very contrasting opinions. Pascal’s statements are in the room, which can also be read in the job descriptions on social networks (and the reactivated transparency article). He assured me that he would make practical videos and otherwise support Alphacool in PR work with influencers and creators in day-to-day business, but not directly in marketing.
Last week, Alphacool told me a completely different view of things and also mentioned a probationary period until the end of the year, but it’s not up to me to judge personnel matters of a foreign company. I shouldn’t even do that. But I always assume that what my counterpart tells me is correct. My dilemma: suddenly there are two supposed truths. But since there seems to be no written contract, the problem is not solvable for me without loss.
In my statement on Saturday, I wrote very clearly that I completely released Pascal from this task. I was already accused of antisocial behavior and shifting my guilt onto suffering third parties. I also told Roman several times on the phone that I don’t kick people who are already down and that everyone should get a second or even third chance. But I also told him that Pascal will no longer do any evaluative reviews or fan tests. But there are enough crafty things that he can certainly and problem-free and under control contribute. So much for social. I even swallow private resentment and various toads because I didn’t like some things. This was before Pascal’s first and Roman’s second video.
Yes, I am very disappointed by many circumstances and can fully understand the reactions of readers and pre-orderers in this regard. But maybe one or the other will now put themselves in my shoes regarding Pascal. Completely executing someone or letting him work heavily supervised in the background is not something you can put to a poll in the forum or an article for voting. I will have to seek advice on this for a long time. As long as the tone is right, we can also do this in the thread of this article.
However, Pascal’s last video and the contradictory statements along with the deletion of threads with Alphacool don’t make the decision any easier. But you just can’t do something like that nowadays. To lump all critics together like that is not possible at all, just as well as removing chat histories. Even if it is a private statement that explicitly only reflects his own point of view, the use of the logos of Alphacool and igor’sLAB is also not excusable. I didn’t even realize it immediately, but I later strongly urged him to put this video on private immediately. I only briefly looked over it and made the mistake of not dissuading him from this idea right away. I completely misjudged that.
It’s his private opinion, which I don’t need to cover up, because it’s his video channel, we don’t have an employment relationship in the traditional sense, and he is fully responsible for himself and his life. However, no matter how emotionally charged one is, one must approach things with more critical ability than he does, and above all, avoid personal attacks. A well-intentioned statement simply must not turn into a shitshow just because one can’t control their adrenaline. Especially not on YouTube. Not because of the possible consequences, but simply out of respect for those who have to watch it (or have to).
This brings us to the third deletion, namely the test of the Core 1. Last week I looked at all the values again in detail and after a long deliberation decided to commission one of the trustworthy freelancers with a new test. This will also take some time, because no one does it as a main job. And so that no false advertising is made with the review until then, the article is also set to private. I generally don’t delete anything, which would also not make any sense. But it was simply the fastest way and was also briefly mentioned in the statement on Saturday. This can be put together later.
What will happen now in detail?
We already discussed this on Saturday, but I will summarize it now. In perspective, we plan to reduce the scope of the flow measurement a bit and instead measure the real cooling performance of a fan on a suitable radiator and maybe even realize this for an air cooler. But first, the backgrounds of the false measurements must be analyzed and especially eliminated. Since there are colleagues who have measured even higher values for the airflow in an open setup, this must also be fully documented and communicated.
The inspection will be taken over by Fritz Hunter. First, new fans of the tested models will be procured, from different manufacturers that are already in the database. Of course not all at once, but some important ones of them. With one of these fans, the reference will be set. Not with respect to the fan, but to the entire measurement system. In other words, Fritz will perform an MSA 1. MSA stands for Measurement System Analysis (MSA). This will first demonstrate the basic measurement system capability. Then two additional fans of the same type and thus a total of three identical fans will be measured 10 times each, thus creating an MSA 3. Which also has to be positive, since otherwise we would not have a valid measurement setup.
As of today, Fritz assumes that the measurement setup will then be valid. Nevertheless, this effort must be made for safety. After the measurement system capability has been demonstrated, there will also be a first article on this. In the sense of an FMEA, Fritz will then not only record all possible errors theoretically, but actually try them out. What this exactly means is stated in the first statement, which is too extensive to lead here.
One of the APEX pre-series fans from Alphacool has also been secured by Fritz at his home, which exactly it is, we have to find out first. At least it has already been newly mounted and thus repaired by Fritz. Since it is beyond our knowledge under what circumstances the fan was exactly measured, a correct retest is not possible in direct form for the time being. Therefore, I have to ask for understanding that we will only measure this again when the measurement setup is validated. Whether we are forced to retest all fans already tested will only emerge through the reprocessing. Since we will use another measurement setup for temperature and plausibility tests in the future, this might also come to us. The measurement setup already outlined must then also first undergo an MSA 1 and MSA 3 at Fritz’s.
Personal Conclusion
This was now a large amount of text, but it was necessary to do justice to all those involved and the problems. Of course, as the operator of the site, I bear the main responsibility for this article and the circumstances why it was not taken offline immediately. I don’t need to sugarcoat that, it’s just the way it is. Hindsight is always wiser. At least I will draw important conclusions from this debacle for myself personally, it can’t be otherwise. It would also bring nothing to look for blame elsewhere. Because you never get to the point. What is appropriate, however, is a request for an apology, which of course does not have to be accepted.
As far as Alphacool is concerned, it is not my job to judge the whole thing, because there they surely know well enough what they should have done differently. Also in dealing with the media and such data. There is certainly still some development potential upwards. That a company is now obviously getting into trouble unnecessarily because of such measurements is completely unnecessary, had the whole thing been marketed better and more thoughtfully. Which brings us back to the euphoria and also my negligence. The marketing slides did not come from me and were not distributed. The dog had already fallen into the ditch. One should have started with damage control earlier. In this respect, I hope for the best for Alphacool, that this will still succeed.
Pascal is a completely different matter. I will not kick him again in this statement, even though especially in some forums and on YouTube a kind of public execution was demanded and some things seem very suspicious to me today. Let me put it this way, he is punished enough by now and certainly needs professional help and good advice, which none of the participants can give him here. In this case, it’s about a person and not an inanimate fan. Please always keep that in mind.
The first statement from Saturday can be read here again:
Personal statement on the test of the Alphacool Apex Stealth (Power) Metal Fan
614 Antworten
Kommentar
Lade neue Kommentare
Veteran
Urgestein
Urgestein
1
Mitglied
Mitglied
Mitglied
Mitglied
Mitglied
Mitglied
Urgestein
Neuling
Mitglied
Mitglied
Urgestein
Urgestein
Veteran
Urgestein
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →